the silver-solution, as is asserted by Lassaigne, I have not yet thoroughly examined; but so much I have ascertained, that a gold-solution of acid reaction is not affected by a gas which produced a distinct reaction in a silver solution. It is, probably, necessary that the gold-solution be perfectly neutral.

§ 31. Which and how many of the above-described tests for ascertaining the nature of the reactions. afforded by Marsh's process have to be made, entirely depends on the number and character of the reactions produced, whether one or several metallic mirrors have been deposited in the reduction-tube, whether or no spots on porcelain have been obtained at the same time, etc.

If a precipitate was produced in the solution of nitrate of silver, the first operation consists in filtering the liquid and adding to the filtrate, very cautiously, ammonia. The formation of a yellow precipitate proves conclusively the presence of arsenic. The addition of ammonia is best performed in such a manner that only a drop at the time is placed cautiously on the surface of the liquid, when two strata of different densities are formed, the upper of an alkaline, and the lower one of an acid reaction. The formation of the precipitate takes place at the plane of contact of the two strata, and by very gradually and carefully mixing the two strata it is not difficult to add just sufficient ammonia to make the precipitate permanent. The precipitate of arsenite of silver thus obtained may be treated with hydrochloric acid; the latter removes the silver, and from the solution the arsenic may precipitated by sulphureted hydrogen; the sulphide

RARY

be

of arsenic may be collected on a filter, dissolved in ammonia, the solution evaporated, and the residue treated with carbonate of soda and cyanide of potassium after the method of Fresenius (v. § 35).

If, besides the metallic mirror, spots on porcelain have been obtained, these must now be made the object of further experiments. A drop of nitric acid of from 1.26 to 1.3 spec. gravity is placed, by means of thin glass-rod, on one of the thickest spots. If, after a few minutes' action, complete solution should not have been effected, a little more acid is added. Application of heat is only required when a weaker acid has been used. In a similar manner a drop of nitrate of silver and ammonia is then added, when the characteristic yellow precipitate is produced, either immediately or after addition of a trace of ammonia, if the spot consisted of arsenic (v. § 27). This decisive experiment must be executed with the greatest care, and requires considerable skill in chemical manipulations. Instead of a solution of nitrate of silver and ammonia, which must be free from an excess of ammonia, a solution of nitrate of silver may be employed, and afterwards ammonia be added, or a glass-rod, moistened with ammonia, be held over the liquid.

If there is reason to apprehend that too much nitric acid was used to effect the solution of the spot, the porcelain may be very gently heated, and the excess of acid be removed by carefully blowing on the drop. In this case, i. e., in the case of an excess of nitric acid, some arsenic acid may have been formed, and on addition of nitrate of silver, a reddish or brownish precipitate may appear; it consists of a

mixture of arsenite and arsenate of silver. To prevent the adhesion of too large a portion of the liquids to the glass-rods, these should be chosen very thin, or better still, be substituted by glass-tubes, very narrowly drawn out at one end; from these the solutions can be made to flow out in very small drops.

The second experiment consists in moistening another spot with a solution of hypochlorite of soda; if it is dissolved, it is an arsenic spot; if not, an antimony spot (v. § 27). If both metals should happen to be present, the arsenic, being the more volatile metal, occupies the outer edge of the spot; and, on adding the reagent, the outer brownish-black part only disappears. If only a small quantity of antimony occurs with the arsenic, the whole of the spot is dissolved (Wackenroder). It is an important feature of this test, that it can be practised on the very thinnest spots, while the silver reaction requires much thicker spots to become apparent.

Another spot, which must not be too thin, may be dissolved in nitric acid, in the manner above described, and the solution mixed with a drop of freshly prepared and very concentrated suphureted hydrogen water. If the spot consisted of arsenic, a pure yellow precipitate of sulphide of arsenic is thrown down, while the precipitate is of an orange color if it consisted of antimony. This experiment may be followed by the important test with sulphide of ammonium (v. § 27).

The examination of the spots, after the methods of Cottereau, Lassaigne, and Slater, should only be performed if there is a sufficient number of them.

§ 32. Considerable importance is attached to the chemist's placing an arsenic mirror, as a corpus delicti, in the hands of the authorities. For this reason, the analyst should always endeavor to obtain two mirrors, the one to be disposed of as mentioned, and the other to serve for further experiments. If only one mirror could be procured, there remains nothing but to divide it. The glass tube is divided by the cut of a sharp file at the middle of the mirror, and broken. For this division, the drawing out of the tube, as described in § 12, offers great conveniences, because, in this arrangement, the mirror is distributed over a larger surface. In using these tubes, it is not even necessary to prepare two mirrors, because the division, if made at the middle of the narrow part, furnishes two very fine mirrors, each of which is usually much more distinct than any one deposited in a common wide tube.

The first experiment to be made with the mirror, consists in holding that end of the tube where the metal is deposited, into a very small flame of a common spirit lamp (v. § 4). If the mirror consists. of arsenic, the flame assumes a bluish-white color, and the alliaceous arsenic odor is evolved; the odor is very distinctly perceived when the heated tube is quickly held under the nose. This peculiar smell is alone sufficient to prove undoubtedly the presence of arsenic. This experiment, also, is best performed with a tube that has been drawn out, and afterwards been cut at the narrow part, because the thin tube becomes sufficiently heated by the very smallest flame, and but a small portion of its contents is lost.

The mirror is now dissolved in nitric acid for further examination, and the solution tested with nitrate of silver. To effect solution, a drop of nitric acid is placed into a small porcelain dish, and cautiously sucked up into the tube containing the mirror. The metal dissolves readily if the acid is of the proper concentration; by blowing into the tube, the drop is brought back into the dish. The tube may be washed out with a drop of water, by proceeding in the same manner. The examination with. nitrate of silver is executed in exactly the same way as prescribed for the examination of spots in § 27. If there is a sufficient quantity of the solution, a portion is tested with sulphureted hydrogen water; and, if there is no lack of material, the behavior of the mirror to hypochlorite of soda may be ascertained (v. § 27).

If the experiments are strictly executed after the manner described, there can no longer remain any doubt as to whether the reactions obtained by the process of Marsh were produced by arsenic, or antimony, or by a mixture of both metals. In the last, and more complicated case, it is very desirable to apply, if possible, the test proposed by Fresenius, as above (§ 29), described; viz.: to treat the metallic mirror in a current of sulphureted hydrogen, and to treat the sulphide thus formed with gaseous hydrochloric acid. If antimony has been found, inquiries should be instituted as to whether antimonials have been adininistered as a medicine, and, if both antimony and arsenic have been detected, it will be necessary to test the antimonial for an admixture of arsenic.